Guidance for Verification of Biological Indicators: Understanding the Acceptance Criteria of D-value and Population Verification

End users of biological indicators (BIs) routinely send samples to third-party laboratories for D-value and population verification. During an initial validation or on an annual revalidation, sterilization cycles are challenged with BIs to demonstrate actual microorganism lethality produced during the sterilization. To present a cycle challenge, these resistant microorganisms must be tough enough to meet specific standard requirements as set out by the Association for the Advancement of Medical Instrumentation (AAMI), the International Organization for Standardization (ISO) or the United States Pharmacopeial Convention (USP) For example, AAMI, ISO, and USP all state that if a BI is used for a validation of steam sterilization at 121 °C, the minimum acceptable D-value or resistance for that BI is 1.5 min (1). These conditions raise the possible need for resistance verification before BI use.

This resistance verification often is done as part of a facility's acceptance criteria for a new lot of BIs coming into the facility and before acceptance and use of the lot for validation work or routine monitoring of the sterilization cycles used at that facility. United States Pharmacopeia (USP) General Chapter <1035> "Biological Indicators for Sterilization" states:

The user may consider conducting a D value assessment before acceptance of the lot [of BIs] (2).

The wording "may consider" is important to note. This phrase means that a D-value assessment is not mandatory, but it may be required according to the individual users or a pharmaceutical company's procedure or protocol for BI-acceptance criteria. USP General Chapter (1035) "Biological Indicators for Sterilization" does not mention under the area of user's responsibility, the testing–result–acceptance criteria. Questions arise as to which criteria to use to accept a lot of BIs if one performs a D-value assessment and what variation allowance is acceptable.

Photodisc

As part of verification, the lot of BIs being tested are expected to meet specific requirements as set out in ISO standards or in the USP as to the accuracy of the label-claimed resistance or D-value and population of the BIs. USP General Chapter (1035) "Biological Indicators for Sterilization" states:

Laboratories that have the capability of performing D value assays could conduct a D value determination using one of the three methods cited in the general test chapter Biological Indicators—Resistance Performance Tests (55) and in the appropriate USP monographs for specific biological indicators (2).

This statement of "methods allowed to be used" refers back to the manufacturers' section in USP General Chapter (55) "Biological Indicators—Resistance Performance Tests" for the test method. If one accepts and performs the test method to be USP-compliant, one can assert that the USP acceptance criteria goes with the method as per USP. The same would be true for BI population verification. If performing a population verification, per USP General Chapter (55) "Biological Indicators—Resistance Performance Tests" (3), one also accepts the acceptance criteria of results not less than 50% or more than 300% of the labeled-certified population.

For example, if one were expecting to comply with USP General Chapter (55) "Biological Indicators—Resistance Performance Tests," the following would apply:

The requirements of the test are met if the determined D value is within 20% of the labeled D value for the selected sterilizing temperature and if the confidence limits of the estimate are within 10% of the determined D value (4).

Once the resistance or D-value has been verified and is within acceptable limits of the label claim and it meets or exceeds minimum acceptance criteria for BI-resistance as indicated in the ISO standard or USP, that particular lot of BIs may now be used for validation work. This BI would be considered an acceptable biological challenge to the sterilization process.

Methods for obtaining a label claim D-value

ISO 11138-1 allows for two of three methods to be used to obtain a label-claim D-value. (5). One may use the most-probable-number method by direct enumeration, a fraction–negative method (such as Limited Spearman–Karber), or assess the D-value accuracy by using the USP survive–kill calculated cycles. Regardless of which of the three methods is used, a resistometer is one piece of equipment that will be needed. A resistometer, also known as a biological indicator evaluator resistometer (BIER), can very quickly and accurately deliver and control very precise sterilization process parameters that are critical to the process.

Various standards developed by the American National Standards Institute (ANSI), AAMI, ISO, and USP have very tight equipment or BIER–vessel operational capabilities that must be met. For example, ANSI-AAMI ST44: 2002 (6) states that with a steam–BIER vessel, the equipment must be capable of the following: hitting the target temperature set point within 10 s or less from the time steam charge occurs; maintaining that set temperature to within ±0.5 °C ; and then at the cycle's end, the postvacuum time to reach atmospheric pressure must be within 10 s or less. ST44: 2002 further specifies that a steam resistometer be capable of measuring such conditions as time (resolution of 00:00:01 s and accuracy within ±00:00:02 s), temperature (resolution of 0.1 °C and accuracy of ± 0.5 °C) and pressure (range of 0 to 60 psia, resolution of 0.1 psia, and accuracy of ±0.5 psia). The duration of the exposure time at a given temperature thus is controlled as exact as possible for both time and temperature.

A steam–BIER vessel is fairly small compared with an autoclave chamber. The small chamber is part of the vessel design that allows for an extremely fast steam charge and rapid increase in chamber temperature. Rather than the normal autoclave come-up time needed to reach a set temperature, a steam–BIER vessel should be very capable of hitting set points for temperature in less than 10 s, and is in many cases, closer to 6 s. This process allows for the very exact exposure time needed to determine BI-resistance characteristics.

If one were performing a fraction–negative method of verifying BI-resistance in such a BIER vessel, one would expose multiple groups of BIs to varying cycle exposure times. For example, if one were attempting to verify the resistance of a particular BI in a steam vessel at 121 °C using the Limited Spearman–Karber fraction–negative method, one could expose 20 BIs per group to various exposure times at 121 °C (7). After exposures, each group of BIs would be transferred to growth medium in an aseptic manner with spore strips and incubated at the appropriate temperature.

One example is to use self-contained ampules. The ampules are purple in color at the beginning of the process. Bacterial spores are suspended in the purple tryptic soy broth (TSB). The color comes from adding a pH indicator. If the spores in the ampule survive exposure and incubation, the ampule will turn yellow as the pH drops as an indication of growth. In this particular example, groups of ampules were run at exposure times from 4.5 up to 12 min. with intervals all at 1.5 min. At the exposure time of 4.5 min, all the BIs within this group survived. At 12 min, all BIs within this group were dead and remained purple. In between these two exposures, the interval exposure of 6, 7.5, 9, and 10.5 min provide fractional results. A fraction of the ampules within a group were dead, and a fraction grew. As the exposure time increased from the initial 4.5-min exposure, more and more BIs were killed. With fractional information such as this, one can use the Limited Spearman–Karber method to determine just how tough the BIs are or what their resistance is to this particular cycle.

Assessing third-party laboratories

It may not appear difficult to get a third-party laboratory to conduct the previously described method for resistance verification in a BIER vessel. Several laboratories offer this service. The price may range from $1500 to $3400 to do a full fraction–negative D-value assessment. The main area of concern is the degree of competency that the laboratory has in performing the D-value assessment. A proper D-value assessment is not as straightforward as one may think. Getting an allowance variation of ±20% may be extremely difficult. Numerous critical elements such as equipment, test methods, recovery media, laboratory techniques, laboratory utensils, and personnel are involved in performing such a test.

Equipment. One issue to consider is when contracting out to a third-party laboratory is whether it has an ANSI–AAMI- or ISO-compliant BIER vessel to perform a D-value assessment. Although this may be obvious, be aware that some vessels may be no more than a well-maintained pressure vessel that does not have the capability to measure and record temperature, pressure, or even time to the accuracy required. Control capabilities and specifications to look at for a steam resistometer are ANSI–AAMI ST44: 2002 (6) and would include:

• time resolution of 00:00.01 s and accuracy of ±00:00:02 s;

• temperature resolution of 0.1 °C and accuracy of 0.5 °C;

• pressure accuracy of 0.5 psia;

• temperature control of 60.5 °C during exposure;

• vacuum level capable of 0.65 psia;

• steam-charge time of 10 s or less from 100 °C to a set temperature.

An instrument should be capable of meeting the afore abbreviated list of equipment specifications. The instrument also should be capable of accurately documenting that the conditions, phases, time, and temperatures that occurred during the cycle actually occurred and were within specifications.

It is unacceptable to have final D-value verification reports where time, temperature, and pressure were handwritten notations on a sheet of paper and not actual resistometer-data printouts. Critical factors such as prevacuuming, come-up time, actual temperature, and pressure during exposure and postvacuuming must be part of the report data. The performance accuracy of the unit is critical. The allowances seem very tight, but widening tolerances only will make verification more difficult. A temperature variation of ±0.5 °C actually allows for a difference of 1.0 °C if one BIER unit at the BI manufacturer is operating on the high side (121.6 °C) and the other BIER unit at the verification laboratory is operating at the low end (120.6 °C). This variation alone may account for at least half of the allowance of ±20% in D-value variation.

The final result report from a contract-testing laboratory should include documentation to provide evidence that all critical cycle parameters were met.

Test methods. ISO 11138-1:1994 Annex E allows for the use of a survivor–curve method or use of fraction–negative methods to initially determine a D-value or BI-resistance. To do a D-value assessment or verification, several test methods can be used. One can perform one of several fraction–negative methods: direct enumeration or survive–kill. In an effort to reduce variables and allow for the closest duplication of the BI manufacturer's procedure for D-value certification, it is important that the same test method be used for D-value verification. USP General Chapter (1035) "Biological Indicators for Sterilization states:

Indicate in the labeling that the stated D value is reproducible only under the exact conditions under which it was determined, that the user would not necessarily obtain the same result....(4).

If the BI manufacturer's certification states that the direct-enumeration method were used to establish that BI's resistance, then direct enumeration (survivor curve) should be the method used for verification. Using different methods for D-value certification and for assessment only allows for the introduction of variables that could affect D-value reproducibility. This strongly applies to situations where one is trying to reproduce the D-value for verification.

If only verification is intended, USP allows for the use of the survive–kill method (8). This technique is rather straightforward and will only involve two cycles being run. One would expose a minimum of 20 replicate samples for the USP procedure or 50 samples for the ISO procedure (9) to each of the USP–calculated survive–kill time cycles. All samples exposed to the survive time must survive the exposure, and all exposed to the kill time must not show growth. Using this method where only two cycles are involved may be much quicker and less expensive than running a full D-value assessment. The survive–kill method calculation is a bit padded, but it still can provide points to check for survivors and lethality. It provides a reference point for resistance. To assist in checking the consistent performance of all units within a particular lot of BIs, survival–kill results will provide that additional information.

Recovery media. Different brands and different lots of both TSB and tryptic soy agar (TSA) may not have the same ability to promote the growth of injured spores. Brands X, Y, and Z of TSA may all perform in a very similar manner for a majority of typical laboratory tasks in obtaining growth for culture streaks and slants with most common laboratory microorganisms. Quantifying the presence of remaining colony-forming units (CFUs) from a BI of injured spores of Geobacillus stearothermophilus is a different matter. Not all brands or lots of recovery media have equal ability to accurately promote the growth of such injured spores. Several articles (10, 11) have been published that demonstrate as much as a full-log difference in the recoverability of injured spores when comparing one brand of TSA with another brand.

The information in Table I (12) shows the variance that is possible in CFU recovery between two different brands of TSA used in population assays of several different lots of BI spore strips. USP General Chapter (55) "Biological Indicators—Resistance Performance Tests" (3) was used to run population assays on six different lots of spore strips containing G. stearothermophilus spores. From the last two dilution tubes in the dilution series while performing the assay, 1-mL aliquots were added to six separate petri dishes.

Table I. Comparative results of colony-forming units

TSA Brand A was added to three of the plates, and TSA Brand B was added to the remaining three plates. The only variable in this exercise was the brand of TSA used for the pour plates. One can see the extreme difference in the spore-recovery ability between the two brands used. A D-value result based upon the direct-enumeration method using Brand A media for several fractional cycles would provide a much different result than one produced using Brand B media. Using Brand A media would give one the impression of a faster log reduction in CFU population, and thus a lower D-value would be calculated than if one used Brand B media.

Laboratory techniques, utensils, and personnel. Several additional variables would make D-value assessment success even more difficult, depending on the laboratory techniques, utensils, and personnel used. Besides recent BIER calibration, utensils such as laboratory pipettes or repeaters have an impact. These utensils and incubators should be in good calibration or within acceptable specifications.

There can be a wide variability in accuracy among technicians doing serial dilutions or plating techniques. A key consideration is whether all BIER cycles are reviewed for compliance to specifications before acceptance. With regard to BI placement into and removal from the BIER–vessel chamber, placement should be consistent from one run to another. The BIs should be removed immediately upon cycle completion and upon cycle initiation and be quickly inserted with little warm-up occurring. In addition, the BI–holding rack should offer protection to the BIs being tested compared with the holding rack used by the manufacturer. All these factors can add up to a test result well outside the accepted ±20% variation.

All these factors considered, a D-value verification may be very difficult to accomplish within acceptable variation. Much discussion has occurred over widening this allowance of ±20% to an even wider margin because of the lack of success in hitting the current range. Variability in skill and technique from one laboratory to another is definitely a contributing factor to the problem of obtaining a D-value verification within allowable limits. This situation also is compounded by equipment or BIER-vessel function differences, calibration, and equipment maintenance.

If one is contracting for a D-value verification, some problems may be in the offing. Even with all the issues aforementioned, some verifications are successful and well within the allowance of ±20% variation. When this situation occurs and is repeated with additional lots of BIs, one can only assume that more than luck is taking place. Both the BI manufacturer and the testing laboratory are running the resistance testing in a very similar manner, and both are paying attention to equipment function, methodology, and calibration–maintenance. They have communicated the necessary factors and duplicate the initial testing. Even if the same laboratory were to do back-to-back D-value testing on the same lot of BIs with the same equipment, result differences will still occur. Small variations may be tolerated, but larger differences must be avoided.

When trying to verify a D-value and problems occur, arrange for the verifying laboratory and the BI manufacturer to communicate. It is possible to get verification for a D-value within ±20%, but some exposures may need to be run again after calibration and equipment are checked. Differences in recovery media, verification method, and equipment calibration all may contribute to a verification problem. This simply should not be a right-or-wrong issue. Communication and a willingness to look at test equipment and procedures can usually solve the differences for interlaboratory variation on verification issues.

Russ Nyberg is the director of technical support and biological indicators at Raven Biological Laboratories, 607 Park Drive, Omaha, NE 68127, tel. 402.593.0781, fax 402.593.0921, russ@ravenlabs.com

References

1. General Chapter (1035), "Biological Indicators for Sterilization," United States Pharmacopeia 28–National Formulary 23 (United States Pharmacopeial Convention, Rockville, MD, 2005), p. 2538.

2. General Chapter (1035), "Biological Indicators for Sterilization," United States Pharmacopeia 28–National Formulary 23 (United States Pharmacopeial Convention, Rockville, MD, 2005), p. 2539.

3. General Chapter (55), "Biological Indicators—Resistance Performance Tests," United States Pharmacopeia 28–National Formulary 23 (United States Pharmacopeial Convention, Rockville, MD, 2005), p. 2244.

4. Official Monograph–Biological, "Biological Indicators for Steam Sterilization, Paper Carrier," United States Pharmacopeia 28–National Formulary 23 (United States Pharmacopeial Convention, Rockville, MD, 2005), p. 259.

5. ISO 11138-1:1994 Annex E, Sterilization of Health Care Products–Biological Indicators, Part 1, General (International Organization for Standardization, Geneva, Switzerland), p. 6.

6. ANSI-AAMI ST44:2002, Performance Requirements for Resistometers (American National Standards Institute, Washington, DC and the Association for the Advancement of Medical Instrumentation, Arlington, VA, 2002).

7. ISO 11138-1:1994 Annex C, Fraction Negative Analysis—MPN Method for Subsequent Determination of D Value by Limited Spearman–Karber Method (International Organization for Standardization, Geneva, Switzerland), p. 10.

8. General Chapter (55), "Biological Indicators—Resistance Performance Tests," United States Pharmacopeia 28–National Formulary 23 (United States Pharmacopeial Convention, Rockville, MD, 2005), p. 2246.

9. ISO11138-1: 1994 Annex X, "Sterilization of Health Care Products—Biological Indicators—Part 1: General (International Organization for Standardization, Geneva, Switzerland) p. 13

10. H. Shintani and J.E. Akers, "On the Cause of Performance Variation of Biological Indicators Used for Sterility Assurance," PDA J.Pharma Sci Technol. 54, p. 332-342.

11. H. Shintani et al., "Validation of D-Value by Different SCD Culture Medium Manufacturer and/or Different SCD Culture Medium Constituent," PDA J.Pharma Sci Technol. 54, p. 6–12.

12. R. Nyberg, "Biological Indicators and Population Verification," Infection Control Today (2000).